open reduction and internal fixation vs. a circular fixator for … · tibia. the combination of...

The PDF of the article you requested follows this cover page.

This is an enhanced PDF from The Journal of Bone and Joint Surgery

2006;88:2613-2623. doi:10.2106/JBJS.E.01416 J Bone Joint Surg Am.The Canadian Orthopaedic Trauma Society

Multicenter, Prospective, Randomized Clinical TrialApplication for Bicondylar Tibial Plateau Fractures. Results of a Open Reduction and Internal Fixation Compared with Circular Fixator

This information is current as of April 5, 2011

Supplementary Material http://www.ejbjs.org/cgi/content/full/88/12/2613/DC1

Reprints and Permissions

Permissions] link. and click on the [Reprints andjbjs.orgarticle, or locate the article citation on

to use material from thisorder reprints or request permissionClick here to

Publisher Information

www.jbjs.org20 Pickering Street, Needham, MA 02492-3157The Journal of Bone and Joint Surgery

http://www.ejbjs.org/cgi/content/full/88/12/2613/DC1

http://www2.ejbjs.org/misc/reprints_perms.dtl

http://www.jbjs.org

http://www.jbjs.org

COPYRIGHT © 2006 BY THE JOURNAL OF BONE AND JOINT SURGERY, INCORPORATED

2613

Open Reduction and Internal Fixation Compared with Circular

Fixator Application for Bicondylar Tibial Plateau Fractures

RESULTS OF A MULTICENTER, PROSPECTIVE, RANDOMIZED CLINICAL TRIAL

BY THE CANADIAN ORTHOPAEDIC TRAUMA SOCIETY

Background: Standard open reduction and internal fixation techniques have been successful in restoring osseousalignment for bicondylar tibial plateau fractures; however, surgical morbidity, especially soft-tissue infection and woundnecrosis, has been reported frequently. For this reason, several investigators have proposed minimally invasive meth-ods of fracture reduction followed by circular external fixation as an alternative approach. To our knowledge, there hasbeen no direct comparison of the two operative approaches.

Methods: We performed a multicenter, prospective, randomized clinical trial in which standard open reduction and inter-nal fixation with medial and lateral plates was compared with percutaneous and/or limited open fixation and applicationof a circular fixator for displaced bicondylar tibial plateau fractures (Schatzker types V and VI and Orthopaedic Trauma As-sociation types C1, C2, and C3). Eighty-three fractures in eighty-two patients were randomized to operative treatment(forty-three fractures were randomized to circular external fixation and forty to open reduction and internal fixation). Follow-up consisted of obtaining a history, physical examination, and radiographs; completion of the Hospital for Special Surgery(HSS) knee score, the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), and the Short Form-36(SF-36) General Health Survey; and recording of complication and reoperation rates.

Results: There were no significant differences between the groups in terms of demographic variables, mechanism of injury,or fracture severity and/or displacement. However, patients in the circular fixator group had less intraoperative blood lossthan those in the open reduction and internal fixation group (213 mL and 544 mL, respectively; p = 0.006) and spent lesstime in the hospital (9.9 days and 23.4 days, respectively; p = 0.024). The quality of osseous reduction was similar in thegroups. There was a trend for patients in the circular fixator group to have superior early outcome in terms of HSS scores atsix months (p = 0.064) and the ability to return to preinjury activities at six months (p = 0.031) and twelve months (p =0.024). These outcomes were not significantly different at two years. There was no difference in total arc of knee motion,and the WOMAC scores at two years after the injury were not significantly different between the groups with regard to thepain (p = 0.923), stiffness (p = 0.604), or function (p = 0.827) categories. The SF-36 scores at two years after the injurywere significantly decreased compared with the controls for both groups (p = 0.001 for the circular fixator group and p =0.014 for the open reduction and internal fixation group), although there was less impairment in the circular fixator group inthe bodily pain category (a score of 46) compared with the open reduction and internal fixation group (a score of 35) (p =0.041). Seven (18%) of the forty patients in the open reduction and internal fixation group had a deep infection. The numberof unplanned repeat surgical interventions, and their severity, was greater in the open reduction and internal fixation group(thirty-seven procedures) compared with the circular fixator group (sixteen procedures) (p = 0.001).

Conclusions: Both techniques provide a satisfactory quality of fracture reduction. Because percutaneous reduction and ap-plication of a circular fixator results in a shorter hospital stay, a marginally faster return of function, and similar clinical out-comes and because the number and severity of complications is much higher with open reduction and internal fixation, webelieve that circular external fixation is an attractive option for these difficult-to-treat fractures. Regardless of treatmentmethod, patients with this injury have substantial residual limb-specific and general health deficits at two years of follow-up.

Level of Evidence: Therapeutic Level I. See Instructions to Authors for a complete description of levels of evidence.

isplaced bicondylar fractures of the proximal end of thetibia involving the articular surface are difficult to treat.Previously, the standard accepted treatment for suchD

A commentary is available with the electronic versions of this arti-cle, on our web site (www.jbjs.org) and on our quarterly CD-ROM (call our subscription department, at 781-449-9780, to order the CD-ROM).

McKee.fm 613 Monday, November 6, 2006 2:00 PM

2614

THE JOU R N A L OF BO N E & JO I N T SU RG ER Y · JB JS .ORG

VO LUM E 88-A · NUM BE R 12 · DECEM BER 2006OP EN RE D UC T I O N AN D IN TER N AL FI X A T I O N V S. A CI RC U LA R FIX A TOR F O R TI B I A L PL A TE A U FR A C T U RE S

fractures was open reduction and internal fixation with platesand screws through an extensile anterior incision1-6. However,while this technique was optimal for fracture visualization, re-duction, and fixation, it required extensive soft-tissue dissectionover the predominantly subcutaneous proximal end of thetibia. The combination of damage from the energy dissipatedthrough the soft tissues from the original injury and the exten-sive surgical dissection led to a high rate of complications in-cluding skin necrosis and infection7-9. Young and Barrackreported infection in seven of eight patients with bicondylar tib-ial plateau fractures treated with medial and lateral buttressplates through an anterior incision, with two patients requiringamputation7. In studies of bicondylar fractures treated similarly,Moore et al. reported deep infection or dehiscence in eight ofeleven patients and Mallik et al. found infection complicatedfour of five such injuries8,9.

As the detrimental effects of excessive dissection of thetenuous soft-tissue envelope and devascularization of the os-seous fragments became apparent, a number of alternativemethods of treatment have been popularized, including percu-taneous reduction and circular frame stabilization, minimallyinvasive techniques and implants, and temporary external fixa-

tion followed by delayed definitive surgery10-20. The advantagesof circular frame fixation (with or without percutaneous lag-screw fixation) include minimal soft-tissue disruption, theability to correct deformity in multiple planes, early knee mo-tion, and the option of spanning the knee in patients with con-comitant ligament injury. Early reports by Stamer et al. (whoreported good or excellent results in sixteen of twenty-threeknees, with three infections) and Watson (who described goodor excellent results in twenty-seven of thirty-one bicondylarfractures, with one infection) confirmed the clinical utility ofthis method10,11. However, there remains doubt as to the qualityof articular reduction with circular fixation, and a direct com-parison with standard reduction techniques has not, to ourknowledge, been performed.

The purpose of our study was to compare the outcomeafter standard open reduction and internal fixation with use ofmedial and lateral plates and that after circular fixation withpercutaneous reduction techniques for patients with displacedbicondylar tibial plateau fractures.

Materials and Methodse conducted a multicenter, randomized, prospective clin-ical trial in which standard open reduction and internal

fixation with medial and lateral plates was compared with cir-cular fixator application with percutaneous and/or limited openreduction techniques for patients with displaced bicondylar tib-ial plateau fractures (types V and VI, according to the system ofSchatzker3,21, and types C1, C2, and C3, according to the system ofthe Orthopaedic Trauma Association [OTA22]). Five university-affiliated level-I trauma centers participated. Sixteen surgeons

W

Fig. 1-A Fig. 1-B

Figs. 1-A through 1-D A thirty-three-year-old man with a displaced bicondylar tibial plateau fracture that was managed with a circular fixator.

Fig. 1-A Preoperative radiograph. Note the intra-articular displacement of the lateral plateau fracture and the varus deformity. Fig. 1-B A pre-

operative computed tomographic scan clarifies the main fracture lines. This step is essential if a percutaneous or limited open reduction and

fixation technique is planned.

McKee.fm Page 2614 Monday, November 6, 2006 2:00 PM

2615



performed the procedures in this series (mean, 5.2 proceduresper surgeon). This study was approved by the Research EthicsBoard at each participating institution.

The primary outcome measure was the Hospital forSpecial Surgery Knee score (HSS score) at two years postop-eratively23. A prestudy power analysis suggested that, to havean 80% chance of detecting a 25% difference between thescores in the groups (with an alpha value of 0.05), approxi-mately forty patients would be required in each group. Sec-ondary outcome measures included the Western Ontario andMcMaster Universities Osteoarthritis Index (WOMAC)24,complications, reoperation, the quality of radiographic reduc-tion, the presence of degenerative osteoarthritis, and scores onthe Short-Form 36 (SF-36) health status questionnaire25.

Patients with a fracture who presented to a participatingcenter were identified by the attending orthopaedic surgeon orresident as being eligible for the study. The patients were thenscreened for suitability, and, if they met study criteria, theywere approached by the research study nurse and informedconsent was obtained for study participation. The patientswere randomized by sequentially numbered, opaque, sealed

envelopes to either standard open reduction and internal fixa-tion with medial and lateral plates (the control group) or per-cutaneous reduction and application of a circular fixator (theexperimental group) in a 1:1 ratio. Patients with a bilateralfracture were randomized once, and the same treatment wasused on both fractures. Although randomization in this typeof trial would ideally be performed intraoperatively, it wasthought that this would necessitate an unacceptable delay inthe logistics of equipment assembly and the construction ofthe circular frame (which most surgeons preferred to do pre-operatively). Therefore, randomization was performed on themorning of the planned operative intervention.

Inclusion CriteriaAll patients had a displaced bicondylar tibial plateau fracture(Schatzker type V and VI and OTA types C1, C2, and C3)20,21

with at least one of the following features: an intra-articularstep or gap of >2 mm, extra-articular translation of >1.0 cmor angulation of >10°, an open fracture, compartment syn-drome requiring fasciotomy, or an associated ligament injuryrequiring repair26-28.

Fig. 1-C

Fig. 1-C An anteroposterior intraoperative radiograph following limited open reduction and percutaneous screw fixation and application of a circular

fixator. The intra-articular fractures have been reduced and the varus deformity has been corrected. Although this fracture pattern required olive wire

placement close to the joint line, which may have resulted in penetration of the capsule, no infection developed. Fig. 1-D Final anteroposterior ra-

diograph made two years postoperatively. While there is some residual deformity of the medial plateau, the knee is stable, the leg is straight, and

the patient had returned to all preinjury activities without pain.

Fig. 1-D


2616



Exclusion CriteriaPatients were excluded if they had any of the following features:a pathologic fracture, definitive surgery more than fourteendays after the injury, a preexisting joint disease (osteoarthritis,inflammatory arthritis, or a prior fracture), a severe systemic ill-ness (active cancer, chemotherapy, insulin-dependent diabetes,renal failure, hemophilia, or a medical contraindication for sur-gery), open growth plates, a vascular injury requiring repair (aGustilo Grade-IIIC fracture29), an age of more than sixty-fiveyears, or a severe head injury (initial Glasgow coma scale scoreof <8) or other neurological condition that would interfere withrehabilitation.

Surgical TechniqueAll procedures in both groups were performed with the patientunder general anesthesia, with prophylactic antibiotic coverageadministered and image-intensifier assistance. Arthroscopic vi-sualization of the reduction was not performed. The timing ofsurgical intervention was left to the discretion of the treatingsurgeon. All procedures were performed by the attending staffsurgeon with resident and/or fellow assistance. In the circularfixator group, the patients were managed with a closed, percu-taneous or limited open reduction of the articular surface fol-lowed by the insertion of percutaneous lag screws to stabilizethe articular fragment(s); no plates were used. To accomplishthis, a variety of techniques, including fracture table traction,ligamentotaxis, use of a femoral distractor, and fragment ma-nipulation with percutaneously inserted elevators or reductionforceps, were used. No bone grafts were used in the circular fix-ator group. An anterior miniarthrotomy to repair an obviousanterior cruciate ligament injury was performed four times inthe circular fixator group. No ligament reconstruction with agraft was performed. Next, a standard Ilizarov circular fixatorwas applied, according to the method of Watson, with a mini-mum of four points of purchase proximally (i.e., three olivewires and a half-pin)11,30,31. Attempts were made to keep theproximal wires at least 1 cm from the joint surface to avoid thecapsular reflection if this was possible32. Then, the shaft was re-duced and fixation was applied through the distal rings of theframe. The circular fixator was composed of three (if the meta-physeal-diaphyseal fracture was very proximal) or four rings inevery case30,31. Compression was applied or angulation was cor-rected at this point (Figs. 1-A through 1-D). The knee was ex-amined following the completion of fixation. If proximalpurchase and/or stability was poor, or a substantial knee liga-ment injury was present, a hinged frame that spanned the kneewas then applied (four knees). It was removed at the discretionof the surgeon (at a mean of 4.5 weeks).

In the open reduction and internal fixation group, stan-dard AO principles of exposure and fixation were used3. A sin-gle anterior incision or combined medial and lateral incisions(at the discretion of the treating surgeon) with arthrotomywere used to perform an open reduction and lag-screw fixa-tion of the articular surface. The menisci and cruciate liga-ments were examined and identified, and soft-tissue injurieswere repaired if possible. Typically, this involved screw and/or

wire and/or suture fixation of avulsed cruciate ligaments withfragments of bone and repair of detached menisci with use ofdrill-holes in bone or suture anchors. Next, plates were ap-plied in every knee medially and laterally to reestablish tibialalignment and buttress the articular repair. No locking plateswere used in this series. Iliac crest bone-grafting was per-formed to support any impacted articular fragments after theywere elevated and secured according to the preference of thesurgeon. Standard wound closure over drains was performed(Figs. 2-A, 2-B, and 2-C).

Open fractures were treated with immediate irrigationand débridement followed by immediate repair or temporaryexternal fixation or splinting until the definitive procedurecould be performed (see below).

Postoperative ManagementIn the open reduction and internal fixation group, knees with-out a ligamentous or meniscal abnormality were placed in aremovable knee immobilizer and allowed early motion.Weight-bearing was restricted for six weeks, and then partialweight-bearing was allowed for an additional six weeks. Par-tial to full weight-bearing was allowed in the circular fixatorgroup (according to the preference of the surgeon) immedi-ately postoperatively. At twelve weeks postoperatively, fullweight-bearing and strengthening were allowed in bothgroups for patients without complications. Postoperative carewas individualized in patients with concomitant knee liga-ment injury or complications.

Follow-upPatients were assessed by the study nurse at the time of inclu-sion in the trial, and they completed outcome forms to docu-ment their preinjury status. In addition to routine clinicalfollow-up, patients were assessed by the study nurse, com-pleted outcome questionnaires, and had radiographs per-formed at six, twelve, and twenty-four months after the injury.

Radiographic AssessmentRadiographic parameters were measured by the study nursecoordinators in each center using a standard measurementtechnique and were recorded on a trial-specific radiographicdata sheet. In difficult knees or when parameters were unclear,assistance was obtained from the principal investigator for thesite or the treating surgeon. Initial measurements and jointparameters at the time of follow-up were made from standard“trauma series” radiographs. Final alignment was calculatedon full-length standing radiographs.

Statistical AnalysisStatistical analysis was performed with use of the SPSS soft-ware package (version 13; SPSS, Chicago, Illinois). A chi-square test was used for categorical variables between the twogroups, such as gender, sex, etc. A traditional Pearson chi-square test was used when statistical conditions were met. AStudent t test was used for continuous variables, such as ageand range of motion. The Fisher exact test was used in cases


2617



when one or more of the expected variables were less than five,and the linear-by-linear association was used when three ormore categorical variables were under consideration (i.e.,mechanism of injury). A p value of <0.05 was considered to besignificant.

Resultsrom November 1998 through May 2003, 116 fractures in114 patients were identified. Twenty-four patients refused

to participate or, after being introduced to the study, insistedon one form of surgical intervention; such patients were fol-

lowed as a separate “nonrandomized” group (see below).Eight patients (two who had preexisting knee osteoarthritis,two who had an ineligible age, one who had vascular injury,and three who had no fixed address) were excluded. Thus,eighty-two patients with eighty-three fractures were random-ized into the study: forty-two patients (forty-three knees) inthe circular fixator group and forty patients (forty knees) inthe open reduction and internal fixation group. Two patientswho were randomized to the circular fixator group underwentan open reduction and internal fixation because of an inabilityto reduce the intra-articular portion of the fracture (both were

F

Fig. 2-A

Figs. 2-A, 2-B, and 2-C A thirty-two-year-old man with a displaced bicondylar tibial plateau fracture that was managed with conventional open reduc-

tion and internal fixation with medial and lateral plates, performed through two incisions. Fig. 2-A Preoperative radiograph. Fig. 2-B Postoperative

radiograph. Fig. 2-C Final radiograph made two years postoperatively. There is mild residual deformity of the lateral plateau, but the leg is clinically

straight, the knee is stable and pain-free, and the patient had returned to preinjury activities.

Fig. 2-B Fig. 2-C

TABLE I Demographic Data on the Patients

ParameterStandard Open Reduction

and Internal Fixation (N = 40)Circular

Fixator (N = 42) P Value

Male gender 23 29 0.278

Workers’ Compensation claims 4 6 0.738

Mechanism of injury* 0.974

Mean age (yr) 43.3 46.2 0.250

Associated systemic injuries 9 12 0.529

Injury Severity Score 9.9 6.6 0.087

*See text for data on mechanisms of injury.


2618



OTA type-C3 fractures). Both of these patients were followedwith use of the intention-to-treat principle (i.e., they re-mained in the circular fixator group). One patient had bilat-eral injury and was randomized to the circular fixator group.

Follow-upSeventy-six patients completed the six-month assessment(two patients, including one who had a bilateral fracture, inthe circular fixator group and four in the open reduction andinternal fixation group were lost to follow-up). One patient inthe open reduction and internal fixation group died of unre-lated causes at nine months postoperatively. Seventy-two pa-tients completed the twelve-month assessment (five in thecircular fixator group and four in the open reduction and in-ternal fixation group were lost to follow-up), and sixty-six pa-tients completed the two-year assessment (nine patients in the

circular fixator group and six in the open reduction and inter-nal fixation group were lost to follow-up). No significant dif-ferences were detected between the groups with respect to thenumber of patients who were lost to follow-up (p = 0.781).

Demographic Data on the PatientsNo significant differences were detected between the groupswith regard to demographic variables (Table I). There was nodifference between the groups (as measured by the linear-by-linear association test) with respect to the mechanism of injury,which included a fall (fifteen patients in the circular fixatorgroup and eight in the open reduction and internal fixationgroup), a motor vehicle collision (eleven and seventeen, respec-tively), a motor vehicle-pedestrian collision (six in each group),a sports-related injury (two and five, respectively), a work-related injury (two and one, respectively), a bicycle accident

TABLE II Injury Characteristics and Associated Injuries




Schatzker type-V fracture 9 9 0.907

Schatzker type-VI fracture 31 34 0.907

OTA fracture type* 0.702

C1 8 12

C2 18 21

C3 14 10

Diaphyseal-metaphyseal translation† (mm) 9 ± 14 9 ± 9 0.849

Diaphyseal-metaphyseal angulation† (deg) 7 ± 8 8 ± 7 0.886

Intra-articular step-gap† (mm) 10 ± 8 9 ± 9 0.817

Fasciotomy (no. of knees) 4 3 0.712

Open injury (no. of knees) 9 5 0.134

Associated injuries (no. of knees)

Neurologic injury 4 2 0.374

Cruciate ligament injury 12 8 0.233

Collateral ligament injury 8 4 0.120

Meniscal injury identified 13 3 0.001

*OTA = Orthopaedic Trauma Association. †The values are given as the mean and the standard deviation.

TABLE III Intraoperative Parameters




Mean total operative time (min) 183 170 0.229

Mean tourniquet time (min) 96 48 0.001

Mean blood loss (mL) 544 213 0.006

Meniscal repair (no. of knees) 12 2 0.008

Cruciate ligament surgery (no. of knees) 10 4 0.436

Hospital stay* (days) 23.4 ± 3.8 9.9 ± 1.6 0.024

*The values are given as the mean and the standard deviation.


2619



(three and one, respectively), and other mechanisms (four andtwo, respectively).

Injury Characteristics and/or Associated InjuriesThere were no significant differences in fracture or injury pat-tern between the groups (Table II). There were nine Schatzkertype-V and thirty-four Schatzker type-VI fractures in the cir-cular fixator group and nine Schatzker type-V and thirty-oneSchatzker type-VI fractures in the open reduction and internalfixation group. According to the OTA classification, there wereeight type-C1 fractures, eighteen type-C2, and fourteen type-C3 fractures in the group managed with open reduction andinternal fixation and twelve type-C1, twenty-one type-C2, andten type-C3 fractures in the circular fixator group. Fracture

displacement was measured after closed reduction and splint-ing and demonstrated no significant differences between thegroups in terms of fracture displacement or angulation, InjurySeverity Score, or associated injuries.

Intraoperative ParametersThe mean total operative time was 170 minutes for the circu-lar fixator group and 183 minutes for the group managed withopen reduction and internal fixation (p = 0.229) (Table III).The mean tourniquet time was significantly less in the circularfixator group (forty-eight minutes) than that in the open re-duction and internal fixation group (ninety-six minutes) (p =0.001). Also, the mean intraoperative blood loss was signifi-cantly less in the circular fixator group (213 mL) than in the

TABLE IV Outcome Data


and Internal Fixation* (N = 40)Circular


Range of motion* (deg)

Flexion 113 ± 32 123 ± 15 0.114

Extension 4 ± 6 3 ± 6 0.499

Total arc of motion 109 ± 33 120 ± 19 0.091

Return to preinjury activity (no. who returned/total no. in group)

6 mo 1/36 8/40 0.031

1 yr 2/35 10/37 0.024

2 yr 4/33 10/33 0.128

Mean HSS knee score†

6 mo 61 72 0.064

1 yr 67 72 0.406

2 yr 68 75 0.307

*The values are given as the mean and the standard deviation. † HSS = Hospital for Special Surgery.

TABLE V Complications and Reoperations


and Internal Fixation* (N = 40)Circular

Fixator* (N = 43) P Value

Incision and drainage 8 2

Split-thickness skin graft 5 2

Screw removal 0 6

Knee manipulation 3 2

Plate removal 8 0

Total knee arthroplasty 2 1

Above-the-knee amputation 1 0

Soft-tissue flap 4 0

Revision open reduction and internal fixation 4 0

Other 2 3

Total 37 16 0.001

*The values are given as the number of reoperations.


2620



open reduction and internal fixation group (544 mL) (p =0.006). There was no difference in the number of patients ineach group who underwent contralateral lower-extremity sur-gery (p = 0.626).

HSS and WOMAC ScoresThere were no differences between the groups with regard tothe mean preinjury HSS knee score (94 for the group man-aged with a circular fixator and 95 for the group managedwith open reduction and internal fixation) (p = 0.724). At sixmonths after the injury, the group managed with a circularfixator demonstrated a trend for a higher HSS score (meanand standard deviation, 72 ± 18) than those in the group man-aged with open reduction and internal fixation (mean, 61 ±23) (p = 0.064) (Table IV). At one year, no difference in themean score was detected between the two groups (72 ± 17 forthe circular fixator group and 67 ± 18 for the open reductionand internal fixation group (p = 0.406). At the two-year follow-up examination, the mean HSS score was 75 ± 19 in the cir-cular fixator group compared with 68 ± 20 in the open reduc-tion and internal fixation group; the difference was notsignificant (p = 0.307) (Table IV). There were no differencesbetween the groups with regard to the mean pain, stiffness, orfunction scores on the WOMAC at six months, one year, ortwo years (see Appendix). However, there were significantlydecreased mean scores at two years after the injury in bothgroups relative to the preinjury score (p < 0.05 for all catego-ries in both groups), confirming significant residual lower-extremity disability.

General Health Status (SF-36) OutcomeThere were no significant differences between the groups withrespect to the general health at baseline. Significant decreaseswere detected in all domains in both groups when the meanpreinjury and two-year follow-up scores were compared (p =0.001 for the circular fixator group and p = 0.014 for the openreduction and internal fixation group). There were no differ-ences between the groups at the two-year follow-up evalua-tion with respect to any SF-36 domain except for bodily pain,which demonstrated significant improvement in the circularfixator group (mean, 46) compared with that in the open re-duction and internal fixation group (mean, 35) (p = 0.041)(see Appendix).

Return to Preinjury Activity LevelRelatively few patients in either group were able to return tonormal preinjury activities. Patients in the circular fixatorgroup were more likely to have returned to the preinjury levelof activity than were those in the open reduction and internalfixation group at the six-month evaluation (eight of fortycompared with one of thirty-six patients, respectively; p =0.031) and at the one-year evaluation (ten of thirty-sevencompared with two of thirty-five patients, respectively; p =0.025). However, this difference was not significant at twoyears (ten of thirty-three patients compared with four ofthirty-three patients, respectively; p = 0.128).

Range of MotionThe group managed with the circular fixator demonstrated atrend toward a superior range of motion of the knee in a num-ber of parameters at the two-year follow-up examinationcompared with the group that had open reduction and inter-nal fixation, but the differences were not significant (i.e., themean total arc of motion was 120° in the circular fixator groupcompared with 109° in the open reduction and internal fixa-tion group, p = 0.091) (Table IV).

Radiographic ParametersThere were no differences in the amounts of initial fracture ar-ticular displacement, diaphyseal-metaphyseal translation, ordiaphyseal-metaphyseal angulation between the groups (seeTable II). Postoperatively, there was no difference between thegroups with respect to the quality of the reduction (p = 0.122for the proportion with anatomic reduction, p = 0.963 forproportion with a step deformity, p = 0.866 for proportionwith a gap deformity, and p = 0.236 for proportion with anabnormal mechanical axis) (see Appendix). At one year post-operatively, there were radiographic signs of osteoarthritis(narrowing of the joint space compared with that in the con-tralateral knee, osteophyte formation, and sharpening of thetibial spines) in ten of the thirty-five knees managed withopen reduction and internal fixation and in fourteen of thirty-seven knees managed with a circular fixator (p = 0.461). Attwo years postoperatively, there were radiographic signs of os-teoarthritis in eleven of thirty-four knees that had open re-duction and internal fixation and in thirteen of thirty-two thathad a circular fixator (p = 0.372).

Hospital StayThe groups were not significantly different with respect to themean number of days between the fracture and the time ofsurgery (3.7 ± 4.4 days for the circular fixator group and 4.2 ±4.6 days for the open reduction and internal fixation group;p = 0.645). However, the mean total length of stay in the hos-pital was significantly shorter for the group that had the circu-lar fixation (9.9 ± 1.6 days) compared with that for the groupmanaged with open reduction and internal fixation (23.4 ±3.8 days) (p = 0.024). This was primarily due to the increasedhospital stay for patients managed with open reduction andinternal fixation who had complications develop and re-quired multiple procedures.

Reoperation and ComplicationsProcedures done prior to the definitive intervention (i.e., irri-gation, débridement, and temporary external fixation for anopen fracture) were not considered a reoperation or a compli-cation. Twenty-seven of the forty-one circular fixators wereremoved in the operating room with use of a brief general orneuroleptic anesthetic; the other fourteen were removed in theclinic and these were considered “planned” procedures—notcomplications or unplanned reoperations (the mean time toremoval of the circular fixator was sixteen weeks). Eighteenpatients managed with open reduction and internal fixation


2621



required thirty-seven unplanned additional surgeries, whereasfifteen patients who had a circular fixator required sixteen un-planned surgeries (p = 0.001). Although it is a subjective find-ing, the procedures in the circular fixator group (pin-trackdébridement, screw removal, and knee manipulation) tendedto be of a lesser magnitude than those done in the open reduc-tion and internal fixation group (above-the-knee amputation,rotational and/or free flaps, incision and drainage with plateremoval, and osteotomy) (Table V).

Nonrandomized PatientsSince the exclusion of a large number of eligible patients canintroduce substantial bias into a clinical trial (through the “se-lection” of ideal patients and the exclusion of “difficult” ones),we recorded and followed potentially eligible patients whowere not randomized (twenty-four who refused to participateand eight who were excluded). No differences were detectedbetween this group (eighteen managed with a circular fixatorand fourteen managed with open reduction and internal fixa-tion) and the randomized group with regard to demographicdata, mechanism of injury, fracture pattern or type, associatedinjuries, Injury Severity Score, or treatment method chosen(p > 0.05 for all parameters).

Discussionlthough standard techniques of open reduction and inter-nal fixation have proven to be routinely successful in re-

storing osseous anatomy1-6, the tenuous soft-tissue coverage andpredilection for high-energy trauma of the proximal end of thetibia has led to a substantial rate of severe complications. Re-ports by Young and Barrack, Moore et al., and Mallik et al.highlighted the risks of conventional plate fixation7-9. While itcould be argued that these series were reported before our de-gree of understanding with regard to the importance of the soft-tissue envelope (and the surgical techniques to preserve it) inthe proximal end of the tibia had improved to its current level,complications still occur. In an article from a leading level-Itrauma center in 2004, Barei et al. described complications as-sociated with plate fixation of high-energy bicondylar tibial pla-teau fractures in eighty-three patients33. While their results werean improvement from those in previous studies, there were stillseven patients (8.4%) who had a deep infection that required anaverage of 3.3 additional procedures for resolution. This was de-spite the use of “modern” techniques including delayed defini-tive surgery (mean, nine days after the injury), small, widelyspaced incisions with minimal soft-tissue dissection, and theuse of low-profile implants.

These complications have prompted the introduction ofalternative techniques for the treatment of these injuries, in-cluding the use of percutaneous reduction techniques and sta-bilization with circular fixator frames. The promising initialresults described by Watson, Buckle et al., and Stamer et al. re-vealed reasonable outcomes with dramatically reduced infec-tion rates (i.e., one deep infection in thirty-one patients)10-12.However, the quality of fracture reduction with the circularfixator technique for bicondylar tibial plateau fractures has

been questioned33. In the present study, we found that therewas no difference in our restoration of the osseous anatomy,either of the articular surface or of the metaphyseal-diaphysealjunction, between the circular fixator group and the open re-duction and internal fixation group. However, it is importantto note that in two of forty-three fractures randomized totreatment with a circular fixator, we were unable to effectivelyreduce the joint surface percutaneously or through a limitedopen incision and were forced to resort to a formal open tech-nique. This does not necessarily mean that the reductionswere perfect in all patients (in either group): there was a con-siderable rate of nonanatomical intra-articular reduction, butthe outcome did not appear to be compromised in these indi-viduals. We were generally successful in restoring overall limbalignment, with only four patients in the open reduction andinternal fixation group and six in the circular fixator groupwho had >5° of metaphyseal-diaphyseal angulation at the timeof the final follow-up. This lends additional support to currentevidence that articular reduction is of less importance andthat a straight leg, a stable knee, and the absence of complica-tions are primarily responsible for optimizing outcome inthese severe bicondylar plateau injuries16-18,20,34.

Another concern that has been raised with regard to thecircular fixator technique is the inability to clearly identify(and potentially repair) intra-articular soft-tissue abnormali-ties. Obviously, these injuries are easier to identify (and re-pair) with open reduction and internal fixation. In our study,meniscal injuries were identified and repaired in twelve (30%)of forty patients managed with open reduction and internalfixation; presumably the rate was similar in the circular fixatorgroup, in which only two repairs were performed throughlimited open incisions. Despite this, no patient in the circularfixator group required secondary surgery for meniscal or cru-ciate abnormalities during the follow-up period of two years,and the clinical outcome scores were similar or superior tothose in the group managed with open reduction and internalfixation. This finding, which is similar to the results fromother series describing the circular fixator technique, bringsinto question the necessity of aggressive identification or re-pair of intra-articular soft-tissue pathological conditions, in-cluding cruciate ligament reconstruction with grafts14-18,20. Wehave found that stiffness, rather than instability, tends to bethe clinically more relevant problem. Also, while prior studieshave pointed out that the rate of associated soft-tissue injuriesis as high as 56%, no study has, to our knowledge, confirmedthat aggressive investigation and treatment (apart from repair-ing obvious soft-tissue lesions evident in the course of fracturefixation) improves clinical outcome26-28,35.

Our study confirmed the significant rate of complicationsseen following open reduction and internal fixation of these dif-ficult fractures, despite the use of “modern” fracture fixation andsoft-tissue handling techniques. The rate of deep infection in theopen reduction and internal fixation group in the present studywas 17%. This is higher than the rate of 8.4% recently reportedby Barei et al., but it is much improved compared with earlierseries7-9,33,36. Reoperations in the open reduction and internal fixa-

A


2622



tion group tended to be of greater severity and included numer-ous soft-tissue procedures for bone and hardware coverage andrevision of fixation. There was a lower reoperation rate in thecircular fixator group, and the unplanned procedures that wererequired were of lesser magnitude (i.e., screw removal).

There were significant residual deficits in both groups, asmeasured by both limb-specific and general patient-orientedoutcome measures. For example, two years following the injury,only 12% of the patients managed with open reduction and in-ternal fixation and 30% of the patients managed with the circu-lar fixator felt that they had returned to all of their preinjuryactivities. There are a number of possible reasons for this. In gen-eral, patient-oriented outcomes are more sensitive to residualdisability following a fracture than are surgeon-based or radio-graphic outcomes, which have been the focus of many previousstudies1-12. The return to “normal” activities may be compro-mised by concomitant injuries, which affected some of the pa-tients in this series. Also, Gaston et al. recently reported on thereturn of objectively measured muscle strength in the leg follow-ing (mostly unicondylar) tibial plateau fractures37. They foundthat only 14% of the patients achieved normal (defined as that ofthe uninjured, contralateral leg) quadriceps strength, while 30%had normal hamstring strength one year following the injury. Itis reasonable to assume that the more severe bicondylar fracturesseen in our trial would exhibit even greater strength deficits withan obvious negative effect on function. This may be the explana-tion for the higher disability measured by the WOMAC functionsubscale (range, 41 to 56) compared with the stiffness (range, 7to 10) or pain (range, 11 to 15) subscales.

There was some evidence (an earlier discharge from thehospital, improved HSS scores at six months, and less bodilypain as measured with the SF-36) that the circular fixatorgroup had a faster recovery from the injury than the groupmanaged with open reduction and internal fixation. Thiscould be due to a number of factors, including the more ag-gressive rehabilitation program in the circular fixator groupwith earlier weight bearing, the lower postoperative complica-tion rate, and the decreased soft-tissue dissection with de-creased pain and swelling. Given the current societal emphasison rapid recovery and return to function, this must be consid-ered an important advantage of the circular fixator techniqueover conventional surgery. In addition, patients managed withthe circular fixator lost significantly less blood, had a shorterhospital stay, and did not require as much operative timecompared with those who had open reduction and internalfixation. While it was not the focus of this study, the circularfixator technique with decreased soft-tissue dissection allowsearlier definitive intervention in patients with a compromisedsoft-tissue envelope compared with current standards of pro-visional external fixation followed by delayed (often by two orthree weeks) definitive surgery33,36. However, an acceptable re-duction may not always be achievable by limited open proce-dures, and a formal open approach may be required.

It is important to emphasize that this study was con-ducted before the advent of locking plates, and it is not a com-parison of the circular fixator and open reduction and internal

fixation with locking plates. Although there are no compara-tive trials that we are aware of, there is some preliminary evi-dence that the use of locking plates for fractures of theproximal end of the tibia may provide superior results com-pared with conventional plate fixation38.

Despite its randomized and prospective nature, our studyhas several weaknesses. There are a number of results in ourtrial that show trends that approach the accepted level of signifi-cance (p < 0.05); for example, the mean 11° increase in the arcof flexion-extension in the circular fixator group at two years (amean of 120° for the circular fixator group compared with 109°for the group managed with open reduction and internal fixa-tion; p = 0.091). It is possible that this represents a “true” (andclinically relevant) improvement in motion that does not reachsignificance because of a sample size that is too small or a stan-dard deviation that is too large or both. This is defined as a po-tential beta error (a failure to confirm a true difference betweenexperimental groups) that could be eliminated by a larger studywith more patients39. Although we used the best available mea-surement tools at the time of the study, a more precise outcomemeasurement tool might also be able to detect a clinically sig-nificant difference in the two techniques. In addition, the finalfollow-up evaluation was only twenty-four months after the in-jury. It is possible that, with time, many of these patients mayhave posttraumatic arthritic change develop in the knee, espe-cially since the articular reduction was imperfect in some ofthem. It is possible, even probable, that the prevalence of degen-erative change may increase with time. However, one would an-ticipate that, since the prevalence and severity of such changewas equal in the groups at two years, the rate of subsequent de-generation would be similar. Also, late reconstructive surgeryfollowing posttraumatic arthritic change after tibial plateaufractures usually requires total knee arthroplasty. In general, ar-throplasty in this setting is technically simpler with a lowercomplication rate if prior incisions, soft-tissue scarring, and im-planted hardware are minimized. This would typically favor thecircular fixator group40. However, the presence of indolent in-fection from old pin tracks in the proximal end of the tibia isalso a concern for future arthroplasty.

In conclusion, both techniques provided a reasonablequality of fracture reduction. Closed reduction and applica-tion of a circular fixator resulted in a shorter hospital stay,fewer and less severe complications, marginally faster returnof function, and similar or superior clinical outcomes com-pared with conventional open reduction and internal fixationwith plates. These benefits are obtained (in the majority of pa-tients) without compromising the quality of fracture reduc-tion. Regardless of treatment method, patients with this injuryhave significant residual limb-specific and general-health defi-cits at two years of follow-up.

AppendixA table showing the WOMAC outcome values and theradiographic reduction parameters at one year and a his-

togram presenting the SF-36 scores at two years are availablewith the electronic versions of this article, on our web site at


2623



jbjs.org (go to the article citation and click on “SupplementaryMaterial”) and on our quarterly CD-ROM (call our subscrip-tion department, at 781-449-9780, to order the CD-ROM).

NOTE: This manuscript was prepared by the Canadian Orthopaedic Trauma Society, c/o MichaelD. McKee, MD, FRCS(C), Division of Orthopaedics, Department of Surgery, St. Michael’s Hospi-tal and the University of Toronto, 55 Queen Street East, Suite 800, Toronto, ON M5C 1R6, Can-ada. E-mail address: [email protected]. Principal Investigators: Michael D. McKeeand Shafique P. Pirani; Lead Investigators (Site): Shafique P. Pirani (Royal Columbian Hospital),Michael D. McKee (St. Michael’s Hospital), David J.G. Stephen (Sunnybrook Health ScienceCenter), Robert Feibel (Ottawa General Hospital), and James N. Powell (Foothills Hospital);Study Design: Shafique P. Pirani, Michael D. McKee, Robert McCormack, Emil H. Schemitsch,David J.G. Stephen, Hans J. Kreder, and Robert Feibel; Data Analysis: Floyd Sekeramayi, MauriZomar, Michael D. McKee, Lisa M. Wild, and Shafique P. Pirani; Radiographic Analysis: FloydSekeramayi, Michael D. McKee, Lisa M. Wild, and Shafique P. Pirani; Manuscript Preparation:Michael D. McKee, Shafique P. Pirani, Robert McCormack, David Stephen, Lisa M. Wild, andMauri Zomar; Patient Enrollment and Assessment: Richard Buckley, James N. Powell, MichaelD. McKee, Emil H. Schemitsch, Robin R. Richards, James P. Waddell, Timothy R. Daniels, EarlR. Bogoch, Hans J. Kreder, Joel Finkelstein, David J.G. Stephen, Terry Axelrod, Shafique P. Pi-rani, Bertrand Perey, H. MacPherson, Thomas Goetz, Robert McCormack, Graham Pate, M.Penner, Robert Feibel, Alan Giachino, David Barei, and John-Paul Veri; Site Coordinators (Site):Lisa M. Wild (St. Michael’s Hospital), Beverly Bulmer and Lisa Conlin (Sunnybrook Health Sci-ence Center), Leslie Elves (Foothills Hospital), Mauri Zomar, Carla Erho, Karyn Moon, and Eliza-

beth Kimmel (Royal Columbian Hospital), and Anna Conway and Heather Hrushowy (OttawaGeneral Hospital).

In support of their research for or preparation of this manuscript, one or more of the authors received grants or outside funding from Smith and Nephew, Ltd., and the Simon Fraser Orthopaedic Fund. None of the au-thors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or non-profit organization with which the authors are affiliated or associated.

doi:10.2106/JBJS.E.01416

References

1. Covall DJ, Fowble CD, Foster TE, Whitelaw GP. Bicondylar tibial plateau frac-tures: principles of treatment. Contemp Orthop. 1994;28:115-22.

2. Agnew SG, Benirschke SK, Mayo KA, Henley MB, Santoro VM. Open reduc-tion and internal fixation of complex tibial plateau fractures. J Orthop Trauma. 1991;5:236.

3. Schatzker J. Fractures of the tibial plateau. In: Schatzker J, Tile M, eds. Ratio-nale of operative fracture care. New York: Springer; 1987. p 279-95.

4. Benirschke SK, Agnew SG, Mayo KA, Santoro VM, Henley MB. Immediate inter-nal fixation of open, complex tibial plateau fractures: treatment by standard pro-tocol. J Orthop Trauma. 1992;6:78-86.

5. Waddell JP, Johnston DW, Neidre A. Fractures of the tibial plateau: a review of ninety-five patients and comparison of treatment methods. J Trauma. 1981;21:376-81.

6. Tscherne H, Lobenhoffer P. Tibial plateau fractures. Management and expected results. Clin Orthop Relat Res. 1993;292:87-100.

7. Young MJ, Barrack RL. Complications of internal fixation of tibial plateau frac-tures. Orthop Rev. 1994;23:149-54.

8. Moore TM, Patzakis MJ, Harvey JP. Tibial plateau fractures: definition, demo-graphics, treatment rationale, and long-term results of closed traction manage-ment or operative reduction. J Orthop Trauma. 1987;1:97-119.

9. Mallik AR, Covall DJ, Whitelaw GP. Internal versus external fixation of bicondylar tibial plateau fractures. Orthop Rev. 1992;21:1433-6.

10. Stamer DT, Schenk R, Staggers B, Aurori K, Aurori B, Behrens FF. Bicondylar tibial plateau fractures treated with a hybrid ring external fixator: a preliminary study. J Orthop Trauma. 1994;8:455-61.

11. Watson JT. High-energy fractures of the tibial plateau. Orthop Clin North Am. 1994,25:723-52.

12. Buckle R, Blake R, Watson JT, Morandi M, Browner BD. Treatment of the complex tibial plateau fractures with the Ilizarov external fixator. J Orthop Trauma. 1993;7:167-8.

13. Mikulak SA, Gold SM, Zinar DM. Small wire external fixation of high energy tibial plateau fractures. Clin Orthop Relat Res. 1998;356:230-8.

14. Morandi MM, Pearse MF. Management of complex plateau tibial fractures with the Ilizarov external fixator. Tech Orthop. 1996;11:125-31.

15. Watson JT, Ripple S, Hoshaw SJ, Fhyrie D. Hybrid external fixation for tibial plateau fractures: clinical and biomechanical correlation. Orthop Clin North Am. 2002;33:199-209, ix.

16. Marsh JL, Smith ST, Do TT. External fixation and limited internal fixation for complex fractures of the tibial plateau. J Bone Joint Surg Am. 1995;77:661-73.

17. Dendrinos GK, Kontos S, Katsenis D, Dalas A. Treatment of high-energy tibial plateau fractures by the Ilizarov circular fixator. J Bone Joint Surg Br. 1996;78:710-7.

18. Watson JT, Coufal C. Treatment of complex lateral plateau fractures using Ilizarov techniques. Clin Orthop Relat Res. 1998;353:97-106.

19. Jennings JE. Arthroscopic management of tibial plateau fractures. Arthros-copy. 1985;1:160-8.

20. Kumar A, Whittle AP. Treatment of complex (Schatzker Type VI) fractures of the tibial plateau with circular wire external fixation: retrospective case review. J Or-thop Trauma. 2000;14:339-44.

21. Schatzker J, McBroom R, Bruce D. The tibial plateau fracture. The Toronto ex-perience 1968–1975. Clin Orthop Relat Res. 1979;138:94-104.

22. Fracture and dislocation compendium. Orthopaedic Trauma Association Com-mittee for Coding and Classification. J Orthop Trauma. 1996;10 Suppl 1: 1-154.

23. Li PL, Zamora J, Bentley G. The results at ten years of the Insall-Burstein II to-tal knee replacement. Clinical, radiological and survivorship studies. J Bone Joint Surg Br. 1999;81:647-53.

24. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically-important patient-relevant outcomes following total hip or knee arthroplasty in osteoarthri-tis. J Orthop Rheumatol. 1988;1:95-108.

25. Ware JE Jr, Snow KK, Kosinski M, Gandek B. SF-36 health survey: manual and in-terpretation guide. Boston: The Health Institute, New England Medical Center; 1993.

26. Bennett WF, Browner B. Tibial plateau fractures: a study of associated soft tissue injuries. J Orthop Trauma. 1994;8:183-8.

27. Delamarter RB, Hohl M, Hopp E. Ligament injuries associated with tibial pla-teau fractures. Clin Orthop Relat Res. 1990;250:226-33.

28. Wilppula E, Bakalim G. Ligamentous tear concomitant with tibial condylar fracture. Acta Orthop Scand. 1972;43:292-300.

29. Gustilo RB, Anderson JT. Prevention of infection in the treatment of one thou-sand and twenty-five open fractures of long bones: retrospective and prospective analyses. J Bone Joint Surg Am. 1976;58:453-8.

30. Fleming B, Paley D, Kristiansen T, Pope M. A biomechanical analysis of the Ilizarov external fixator. Clin Orthop Relat Res. 1989;241:95-105.

31. Khalily C, Voor MJ, Seligson D. Fracture site motion with Ilizarov and “hybrid” external fixation. J Orthop Trauma. 1998;12:21-6.

32. DeCoster TA, Crawford MK, Kraut MA. Safe extracapsular placement of proxi-mal tibia transfixation pins. J Orthop Trauma. 1999;13:236-40.

33. Barei DP, Nork SE, Mills WJ, Henley MB, Benirschke SK. Complications asso-ciated with internal fixation of high-energy bicondylar tibial plateau fractures utiliz-ing a two-incision technique. J Orthop Trauma. 2004;18:649-57.

34. Katsensis D, Athanasiou V, Megas P, Tyllianakis M, Lambiris E. Minimal internal fixation augmented by small wire transfixion frames for high-energy tibial plateau frac-tures. J Orthop Trauma. 2005;19:241-8. Erratum in: J Orthop Trauma. 2006;20:69.

35. Holt MD, Williams LA, Dent CM. MRI in the management of tibial plateau frac-tures. Injury. 1995;26:595-9.

36. Oh JK, Oh CW, Jeon IH, Kim SJ, Kyung HS, Park IH, Kim PT, Ihn JC. Percutane-ous plate stabilization of proximal tibial fractures. J Trauma. 2005;59:431-7.

37. Gaston P, Will EM, Keating JF. Recovery of knee function following fracture of the tibial plateau. J Bone Joint Surg Br. 2005;87:1233-6.

38. Cole PA, Zlowodzki M, Kregor PJ. Less Invasive Stabilization System (LISS) for fractures of the proximal tibia: indications, surgical technique and preliminary re-sults of the UMC Clinical Trial. Injury. 2003;34 Suppl 1:16-29.

39. Lochner HV, Bhandari M, Tornetta P 3rd. Type-II error rates (beta errors) of randomized trials in orthopaedic trauma. J Bone Joint Surg Am. 2001;83:1650-5.

40. Weiss NG, Parvizi J, Trousdale RT, Bryce RD, Lewallen DG. Total knee arthro-plasty in patients with a prior fracture of the tibial plateau. J Bone Joint Surg Am. 2003;85:218-21.


open reduction and internal fixation vs. a circular fixator for … · tibia. the combination of...

Documents